Ultra Dense Networks, UDNs are generally regarded as the next generation communication system following the 3rd Generation Partnership Project Long Term Evolution (3GPP LTE) system. They are expected to provide data rates up to 10 Gbit/s and even higher, which might meet user requirements for increasing data traffic.
MMW is one of candidate techniques for the next generation communication system (i.e. UDN) due to its high frequencies and huge license-free bandwidth at these frequencies. Compared with technologies operating at lower frequencies, however, MMW may suffer more severe signal attenuation and higher total noise power which result in very short transmission distances. In order to solve this problem, an antenna-array beamforming technology is used which is a promising technology and can enhance intensity of signals or increase transmission distances significantly.
There are two main schemes about the antenna-array beamforming technology. One is a digital beamforming scheme, and the other is an analog beamforming scheme. With respect to both of the two beamforming schemes, weighting operations are generally performed to signals which are to be transmitted from an antenna-array of Access Node (AN) and received by antennas of an antenna-array of a UE (User Equipment), or vice versa. Different weighting operations form different beamforming patterns which thus result in different effects. The digital beamforming scheme performs weighting operations to signals in digital domain and can thus achieve an effect which looks like that signals from all direction can be received simultaneously by an antenna-array of a UE. However, the analog beamforming scheme performs weighting operations to signals in analog domain and can thus achieve an effect which looks like that signals from only one direction can be received at a certain time by the antenna-array of the UE.
In the digital beamforming scheme, since the weighting operation is done in digital domain, each antenna branch needs a dedicated Analog-to-Digital Converters (ADC). But in the analog beamforming scheme, since the weighting operation is implemented in analog domain before a common ADC, an antenna-specific ADC can be avoided. Since complexity and cost of the analog beamforming scheme are lower than the digital beamforming scheme, the analog beamforming scheme is preferably used for UEs or ANs in a MMW-based network. In the current 3GPP LTE network, however, the digital beamforming scheme is used for UEs or ANs. Since, as described above, the effect achieved by the digital beamforming scheme is different from that achieved by the analog beamforming scheme, mechanisms for UEs or ANs using the digital beamforming scheme may not work for UEs or ANs using the analog beamforming scheme.
One of the mechanisms is intra-frequency measurement performed during cell search. In the 3GPP LTE network, intra-frequency measurement refers to a situation in which a neighbouring AN which is working at the same frequency as a served UE being currently serviced by a serving AN, is detected by the served UE. Since the neighbouring AN, the serving AN and the UE work at the same frequency and the digital beamforming scheme is used by the UE, the UE can receive, in terms of actual effect, a signal from the neighbouring AN while performing data communication with the serving AN at the same time. Therefore, in the 3GPP LTE network, measurement to the neighbouring AN and data communication with the serving AN are arranged to be performed simultaneously.
With respect to UEs using the analog beamforming scheme, however, the above mechanism cannot work since the UEs can receive, in terms of actual effect, signals in only one direction at a certain time. When a served UE is performing data communication with a serving AN, measurement to a neighbouring AN which is working at the same frequency as the served UE cannot be performed by the served UE. Therefore, if the cell searching mechanism of the 3GPP LTE network is directly used in a MMW-based network, those neighbouring ANs which are working at the same frequency as the served UE cannot be detected.
Generally speaking, a system signal, such as a reference signal or a synchronization signal should be periodically transmitted by an AN to its served UEs, so that the served UEs can know information about the AN. Furthermore, in UDN networks, it is more possible for ANs to use longer system signal period in order to save power consumption. Therefore, a UE will take much more time to capture a system signal of a neighbouring AN if it starts measurement to the neighbouring AN randomly.